Extent and pattern of DNA methylation under control conditions and salinity stress

Fifteen primer combinations (Table S1) were used to assay cytosine methylation at 5’-CCGG-3’ sequences in two rapeseed cultivars found to be either tolerant (Exagone) or sensitive (Toccata) to salt stress during different seedling developmental stages. Samples were collected from plants grown under stress and control conditions at 4, 7, and 14 days after sowing (DAS). Moreover, samples after recovery from salinity stress (samples subjected to salinity stress until 14 DAS and then grown under non-stress conditions for one or three days) were collected at 15 and 17 DAS for inclusion in the analysis (Figure 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Stages of collection of seedlings.

Seeds were sterilised in 0.1% NaClO, incubated in distilled water (0 mmol/L), and placed on moistened filter paper. Four days after sowing (4 DAS), seedlings (A) were split and grown in boxes containing distilled water (control; B, D, F, H, J, L, O) or NaCl solution (100 mmol/L; C, E, G, I, K). Moreover, at 14 DAS samples grown under stressed conditions (K) were split: half were kept under stressed conditions (N and Q) and the other half were grown under control conditions (recovery; M and P). All samples were collected at 8 am, except for samples indicated with the letter e (evening) which were collected at 8 pm. Samples 4e, 5 and 5e were used for qRT-PCR and not for MSAP analysis.

https://doi.org/10.1371/journal.pone.0075597.g001

A total of 772 and 783 clear and reproducible bands (Figure S1) were amplified from Exagone and Toccata samples, respectively. Under the control conditions, the total methylation of CCGG sequences averaged 46.58% in Exagone and 41.02% in Toccata. The extent of DNA methylation ranged from 45.85% (7 DAS) to 47.67% (4 DAS) in Exagone and from 39.71% (14 DAS) to 43.55% (4 DAS) in Toccata samples (Table 1). Salinity stress decreased the percentage of total methylated bands in tolerant Exagone (avg of 38.27%), but increased it in sensitive Toccata (avg of 48.66%), when levels of DNA methylation were compared with those in the respective unstressed control. These findings indicated opposite effects of salinity stress on DNA methylation in genotypes of B. napus with different levels of tolerance to salinity stress (Table S2). The fully methylated loci were always more abundant than the hemi-methylated ones (Table 1).

In Exagone, level of methylation (avg of 49.67%) and banding pattern after recovery from stressed conditions resembled those observed under control conditions. However, the relative abundances of fully and hemi-methylated bands changed during the recovery period. In fact, whereas there was an evident increase in fully methylated bands in samples after recovery from salinity stress (avg of 46.69%) compared with the proportion in salinity-stressed ones (avg of 35.75%), the levels of hemi-methylated bands remained similar (avg of 2.98% and 2.52% in samples after recovery and salinity-stressed samples, respectively). Methylation of Toccata in recovered samples remained at a level similar to that of salinity-stressed ones (avg of 50.25% vs. 48.66%), even though the relative abundance of fully and hemi-methylated bands changed. In fact, there was evidence of a decrease in the number of hemi-methylated bands (from an avg of 5.16% to 3.44%) and an increase of fully methylated bands (from an avg of 43.48% to 46.8%) in recovered samples compared with salinity-stressed ones.

Salinity-induced changes in the level of methylation in genotypes of rapeseed that differ in their tolerance of salinity stress

Consistent with the approach used by Karan et al. [32], all possible banding patterns between control and salinity stress in tolerant and sensitive rapeseed genotypes were compared to identify the changes in cytosine methylation patterns under salinity stress. Sixteen banding patterns were apparent from the MSAP analysis (Table 2, Figure S1). Patterns A–D represent monomorphic classes in which the methylation pattern is the same following either the control or the salinity-stress treatments. Patterns E–J are indicative of cytosine demethylation, whereas possible cytosine methylation events induced by salt stress are represented by patterns K–P [32].

Out of 772 and 783 bands, 80.3% and 86.1% of CCGG sites remained unchanged after the imposition of salinity stress on the tolerant (Exagone) and sensitive (Toccata) cultivars, respectively (Table 2). The percentages of demethylated sites under conditions of salt stress were 15.7% and 4.1% in Exagone and Toccata, respectively (Table 2). This indicated relatively more DNA demethylation events in the salt-tolerant genotype than in the salt-sensitive one (Table 2). The percentages of sites methylated under salt stress were 4% and 9.8% in tolerant and sensitive genotypes, respectively (Table 2). This indicates relatively more DNA methylation events in salt-stressed sensitive than in salt-stressed tolerant genotypes (Table 2). The test of independence between three different methylation patterns and salt treatments, control and salt stress conditions was carried out using chi-square test (Table S2). The chi-square test results suggest association between stress conditions and the level of methylation both in Exagone and Toccata (i.e. 47.66 and 27.37 in Exagone and 58.99 and 30.69 in Toccata at 14 DAS, P<0.0001; Table S2).

Alteration of DNA methylation pattern under salinity stress and after subsequent recovery

To identify the DNA methylation changes (i.e. demethylation or methylation under salinity stress and subsequent recovery), we classified all differentially methylated DNA fragments into various classes. As indicated in Table 3, the a, b, and c classes included bands with DNA demethylation induced by salinity; the d, e, and f classes comprised methylated DNA fragments induced by salinity stress; and the g and h classes included DNA fragments for which salinity stress had no effect on their methylation status. As many as 71.9% and 84.4% of the demethylated DNA bands were remethylated (class a), whereas only 26.4% and 15.6% remained hypomethylated (class b) after recovery in Exagone and Toccata, respectively. Only a few bands (2 in Exagone and none in Toccata) were found to belong to class c (demethylated by salinity but re-methylated with a different pattern after recovery).

Different behaviour was observed in Exagone and Toccata in terms of the methylated DNA fragments induced by salinity stress. For the salt-tolerant Exagone cultivar, recovery affected the methylation status of only a few of the fragments subject to salinity-induced DNA methylation (12.9%, class d), whereas the vast majority remained unchanged after recovery from salinity stress (67.8%, class e), with only six bands of class f (methylated by salinity, but demethylated with a different pattern after recovery). After recovery, the 77 salinity-induced DNA methylated fragments of the salt-sensitive Toccata cultivar were almost equally distributed between the three classes (36.6%, 28.6%, and 35% for classes d, e, and f, respectively).

Sequencing and bioinformatics analysis of methylated DNA fragments

A total of 19 bands (avg size of 226 bp, ranging from 89 to 363 bp) were excised from acrylamide gels, cloned, and sequenced. Table S3 shows the methylation pattern of all sequenced fragments collected from Exagone and Toccata samples during seedling development under salt stress and control conditions.

Searches for similarities with known genes as well as with genome sequences were undertaken by investigations of the Brassica (B. rapa and B. oleracea) genomes database [45] and the Arabidopsis TAIR repository [46] using the BLAST alignment tool [47].

Nine sequences showed at least one significant alignment (e-value lower than 0.05) in either of the two Brassica genomes or the A. thaliana (Table 4) genome. In particular, five fragments were significantly associated with B. rapa chromosomes (e-value lower than 0.001), and were also confirmed by alignments with the B. oleracea genome sequence. Among these fragments, only Bn_01 overlaps with a coding sequence, which encodes CYP86A8 (Lacerata, LCR), a member of the CYP86A subfamily of cytochrome p450 enzymes (Table 4). In particular Bn_01 is located 181 bp from the start codon. The Bn_02 sequence is located between the Bra027033 (similar to At1g62600, which encodes a flavin-binding monooxygenase protein) and Bra027034 genes (similar to At3g02340, which encodes a RING/U-box superfamily protein) (Table 4 and Table S4). The Bn_03 sequence is located between the Bra034912 (similar to At4g12760, which encodes an unknown protein) and Bra034913 genes (similar to At1g35490 encoding a bZIP family transcription factor) (Table 4 and Table S4). The Bn_05 sequence is located between the Bra023103 (similar to At2g37170, PIP2, Plasma Membrane Intrinsic Protein 2, which is known to be involved in the response to salt stress) and Bra023104 genes (similar to At3g53430, which encodes ribosomal protein L11, involved in RNA methylation) (Table 4 and Table S4). The Bn_06 sequence is located between Bra038073 (similar to At4g15740, which encodes a calcium-dependent lipid-binding protein) and Bra038074 genes (Table 4 and Table S4).

Moreover, one fragment was associated with the B. rapa chloroplast genome (Bn_04) and the last three fragments were only supported by tBLASTx results with either the Brassica genomes or the Arabidopsis genome. Specifically, Bn_07 and Bn_08 showed similarities to the B. rapa genome sequence, whereas the Bn_09 sequence overlapped (804 bp from the start codon) with the A. thaliana gene, which encodes TPS4 (Trehalose Phosphatase/Synthase 4), a protein involved in trehalose biosynthesis.

Expression analysis

Expression analysis of the two methylated sequences similar to Lacerata (LCR, CYP86A8, Bn_01) and trehalose-6-phosphatase synthase S4 (TPS4, Bn_09) was carried out by quantitative RT-PCR on root and shoot samples collected from the two contrasting genotypes under control, salinity stress, and recovery conditions at various seedling developmental stages (from 4 to 17 days after germination, Figure 1). Efficient qRT-PCR amplifications with a single amplicon were achieved for both genes.

Exagone samples grown under non-stressed conditions showed expression of the LCR gene at stages 4 and 4e (Figure 2 A and C, light-blue bars) that was much higher than that seen for Toccata (Figure 2 B and D, light-blue bars) in both shoots and roots (P<0.01). Observed differences were statistically significant (Student’s t-test, Table S5). In all other shoot stages, the expression of LCR remained higher in Exagone samples, although the differences were smaller (but always statistically significant at at least P<0.01). Not all the differences in LCR expression in the equivalent root stages reached the statistical significance (Table S5). More importantly, LCR, which was methylated under salinity stress in the Exagone cultivar, was downregulated after exposure to salinity stress in both shoots and roots of the tolerant Exagone cultivar from stage 4e to stage 17 (Figure 2 A and C, red bars; P<0.01, Table S6), whereas its expression remained unchanged in both shoots and roots of the sensitive Toccata cultivar (Figure 2 B and D, red bars; Table S6). This expression confirms the pattern of methylation and demethylation observed using MSAP (methylated only in Exagone and no methylation changes in Toccata, Table S3). The expression of LCR in Exagone samples after recovery from salinity stress rapidly returned to levels comparable to those of samples grown under non-stressed conditions (Figure 2 A and C, blue bars; Table S6).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Expression profiles of Lacerata (LCR) in rapeseed after exposure to salinity stress.

Salinity-stressed shoots (A and B) and roots (C and D) of two rapeseed cultivars, Exagone (salt-tolerant, A and C) and Toccata (salt-sensitive, B and D), were used to quantify the expression of LCR relative to control. Real-time PCR analysis was performed using gene-specific primers. The expression of LCR was normalised relative to the expression of internal control genes, which encode glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and elongation factor 1-alpha (EF-1- α. Error bars indicate the standard error calculated using a 95% confidence interval. Relative expression values (fold change) are on a logarithm base 2 scale.

https://doi.org/10.1371/journal.pone.0075597.g002

Under control conditions, the overall expression levels of the TPS4 gene in Exagone (Figure 3 A and C, light-blue bars) were lower than those for Toccata samples (Figure 3 B and D, light-blue bars) in both shoots and roots at almost all stages (P<0.01, Table S5). Interestingly, in Exagone samples, the TPS4 gene was upregulated under salinity stress from stage 5e to stage 17 (Figure 3 A and C, red bars), whereas in Toccata samples, it was downregulated (Figure 3 B and D, red bars) in both shoots and roots compared with the levels in samples grown under control conditions (P<0.01, Table S6). This expression confirms the pattern of methylation and demethylation observed using MSAP (demethylated in Exagone and methylated in Toccata, Table S3). The expression of TPS4 in both Exagone and Toccata recovered samples remained at levels comparable to those of samples grown under stressed conditions for a while, and then returned to levels comparable to those of samples grown under non-stressed conditions at stage 17 (Figure 3, blue bars; Table S6).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Expression profiles of TPS4 in rapeseed subjected to salinity stress.

Salinity-stressed shoots (A and B) and roots (C and D) of two rapeseed cultivars, Exagone (salt-tolerant) and Toccata (salt-sensitive), were used to quantify levels of TPS4 transcripts relative to those in unstressed control plants. Real-time PCR analysis was performed using gene-specific primers. The expression of the TPS4 gene was normalised relative to the expression of internal control genes, which encode glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and elongation factor 1-alpha (EF-1-α. Error bars indicate the standard error calculated using a 95% confidence interval. Relative expression values (fold change) are on a logarithm base 2 scale.

https://doi.org/10.1371/journal.pone.0075597.g003

In order to test the correlation between the expression level of TPS4 and the accumulation of trehalose, total soluble sugar was extracted from Exagone leaves and Toccata roots and quantified. We chose these tissues because they were the tissues where the expression of TPS4 differed significantly between stressed and control samples. In shoot samples of Exagone the level of trehalose in stressed samples was 2.3 fold that of control samples both at 7 DAS and 14 DAS (Student’s t-test 18.88 and 20.84; P-value 4.6 e-05 and 3.1 e-05, respectively; Table S7). In root samples of Toccata, instead, the level of trehalose did not statistically varied between stressed and control samples (Student’s t-test 0.47 and 0.38; P-value 0.65 and 0.72, respectively; Table S7).

Bisulfite sequencing

Bisulfite sequencing of LCR and TPS4 loci was conduct to assess the cytosine methylation status. The bisulfite nucleotide sequences from top strand of these loci were similar to their corresponding loci from Brassica (B. rapa and B. oleracea) genomes database [45] and the Arabidopsis TAIR repository [46] and most of the methylated cytosines belonged to CG types (Figures S2 and S3). Alignment of LCR (Figure S2) sequence with the bisulfite treated DNA sequences from shoots and roots of Exagone and Toccata under non-stress and salinity stress conditions at 7 and 14 DAS revealed that methylated cytosines present under non-stressing conditions at position 4 and 37 did not change even under salinity stress both in Exagone and Toccata. We found genotype-specific methylations such as the one of cytosine at position 49, which is methylated in both tissues, and both conditions in Exagone but un-methylated in Toccata. Moreover, cytosines at position 145 and 162 are methylated under both conditions in roots of Exagone and Toccata, respectively, but un-methylated in shoots of both cultivars. Finally, and more importantly, cytosine at position 58 (involved in the HpaII/MspI restriction site) is methylated only in Exagone under salt stress.

In case of TPS4 (Figure S3) cytosines at 181 and 219 positions were methylated under both conditions and both tissues in Toccata and Exagone, respectively. Cytosine at position 204 was methylated under non-stress and salt stress only in Toccata roots. At position 167 cytosine (belonging to the HpaII/MspI restriction site) was methylated under control conditions in Exagone in both tissues. The methylation status changed in shoots and roots under salt stress. The opposite situation was observed for Toccata. In fact in this genotype, cytosine at position 167 showed to be un-methylated under control conditions and was methylated as consequence of salinity stress.

Plant material and growth conditions

Seeds of two rapeseed cultivars, reported to be tolerant (Exagone, Monsanto) and sensitive (Toccata, Maisadour Semences) to salt and drought stress during germination [43], were sterilised in 0.1% NaClO (v/v) and incubated in distilled water (0 mmol/L). Seven samples each of 70 seeds for each cultivar were placed on moistened Whatman No. 1 filter paper in 9-cm-diameter Petri dishes, and incubated under darkness conditions at 20 °C in a controlled chamber. Four days after sowing (i.e. around 1 day after germination), seedlings were transferred to Plexiglas boxes with holes drilled in the lid and walls (21 holes of 3 mm diameter) to perform the slant test [73]. Seedlings were placed 1 cm apart in a horizontal row along the long axis of a 10 × 16 cm rectangular Whatman No. 1 chromatography paper held on a 10 × 18.5 cm clear plastic plate 3 mm thick. The plate was held at an angle of 20° from the vertical by a slotted rack in the base of the 20 × 30 cm seed tray. The paper was kept moist by dipping it into the solution in the tray. Each plate held as many as 10 seedlings, and each tray held 10 plates. Seedlings from each cultivar were placed in boxes containing distilled water (0 mmol/L) or a NaCl solution (100 mmol/L), both added with Flory 9 (Agrimport), a liquid fertiliser for hydroponics. The boxes for the slant test were placed in a growing room for 17 days at day/night temperature of 25/15 °C with a photoperiod of 14 h light (irradiance: 250 µmol m⁻² s⁻¹) and 10 h darkness. A randomised block design with three replicates per treatment (cultivar × growing conditions) was adopted. Stressed and non-stressed plants were retrieved from the plates at specific intervals (Figure 1). After 14 days, some of the samples grown under stress conditions were transferred to non-stressful conditions for 3 days to assess whether exposure of the plants to non-stressful conditions enabled the plants to recover the “normal” patterns of DNA methylation and gene expression (hereafter named “recovered samples”). Samples were collected in triplicate (biological replicates), snap frozen in liquid nitrogen, and stored at -80 °C. Collection was carried out at 8am for samples 4, 5, 7, 14, 15 and 17 (where the number means the days after sowing) and at 8pm for samples 4e, and 5e (Figure 1).

DNA extraction and methylation sensitive amplified polymorphism (MSAP) analysis

Whole seedling genomic DNA was extracted from triplicate samples collected at 4, 7, 14, 15, and 17 (Figure 1) using the GenElute^TM Plant Genomic DNA Miniprep Kit (Sigma-Aldrich). The integrity and size of genomic DNA was checked using 1% agarose gel electrophoresis prior to its use for MSAP analysis. To detect MSAPs, two digestion/ligation reactions were carried out simultaneously. In the first reaction, 500 ng of the genomic DNA was added to a 45 µl mix containing 5 units EcoRI, 5 units HpaII (New England Biolabs), 1X Restriction-Ligation buffer (1X One Phor All, Amersham Pharmacia, added with 0.1 M DTT and 250 ng BSA), 50 pmol HpaII adapter, 50 pmol EcoRI adapter, 10 mmol ATP, and 1 unit T4 DNA Ligase (Invitrogen). The mixture was incubated at 37 °C for 4 h. The reaction was stopped by incubation at 65 °C for 10 min and then diluted 10 times in 0.1X TE (1 mM Tris–HCl, 0.1 mM EDTA, pH 8). The second digestion/ligation reaction was carried out in the same way, except that MspI was used instead of HpaII.

Two consecutive PCRs were performed to selectively amplify the EcoRI-HpaII and EcoRI-MspI fragments. The pre-selective amplification was performed using 5 µL of the above-mentioned diluted mixture, which was added to a 15 µL mix containing 50 ng of EcoRI+C or EcoRI+A primer (Table S1), 50 ng of HpaII+T/MspI+T or HpaII+A/MspI+A primer, 1X PCR Buffer (Invitrogen), 5 mM dNTPs (Invitrogen), and 1 unit of Taq DNA polymerase (Invitrogen). The PCR cycling conditions were 1 cycle of 45 s at 94 °C, 30 s at 65°C, and 1 min at 72°C, followed by a touch-down profile (13 cycles with -0.7 °C decrease in temperature for the annealing step in each successive cycle), followed by 18 cycles with annealing at 55.9 °C, and finally by an extension cycle of 10 min at 72 °C.

Selective amplifications of the diluted pre-selective amplified products was carried out using a total of 15 primer combinations (Table S1). For each reaction, 5 µL of 1:10 diluted pre-selective amplified samples was added to the following selective amplification mix: 50 ng of EcoRI+2 or EcoRI+3 primer, 50 ng of HpaII/MspI+3 primer (Table S1), 1X PCR Buffer (Invitrogen), 5 mM of dNTPs (Invitrogen), and 1 unit of Taq DNA polymerase (Invitrogen) in a final volume of 20 µL. The temperature profile for selective amplification PCR reaction was the same as that used for the pre-selection step. One µL of each amplified sample was added to 10 µL of formamide and to 0.5 µL of size standard (Genescan ROX 500, Applied Biosystems). After denaturation (94 °C for 5 min) amplified fragments were separated with the ABI 3130xl Genetic Analyzer sequencer.

Amplified fragments were divided into four types based on the presence or absence of bands, which resulted from the differential sensitivity of the fragments to digestion by MspI and HpaII. Type I represents the presence of bands in both enzymes combinations i.e. EcoRI/HpaII and EcoRI/MspI, type II bands appeared only in EcoRI/HpaII but not in the EcoRI/MspI, type III generated bands in EcoRI/MspI but not in the EcoRI/HpaII, and type IV represents the absence of bands following both enzyme combinations. Type II indicates the hemimethylated state of DNA that results from methylation in one DNA strand but not in its complementary strand [33]. Type III represents the case of full CG (internal cytosine) methylation, whereas type IV is the case of full methylation at both cytosines.

Silver staining and DNA sequences of salt-stress-related fragments

Some samples, which were chosen on the basis of interesting polymorphisms, were run on acrylamide gels and silver stained with the aim of isolating and sequencing the selected bands. To do so, 2 µL of selected samples were added to 1X formamide dye (98% formamide, 10mM EDTA, 0.01% w/v bromophenol blue and 0.01% w/v xylene cyanol) and denatured (heated at 95 °C for 3 min). After denaturation, samples were immediately transferred to iced water for 2 min, loaded onto a 6% denaturing polyacrylamide gel, and run for 3–4 h at 40–55 W at 55 °C. Gels were then silver stained. The gels were fixed by incubation in 10% acetic acid for 20 minutes, washed three times with ultrapure water for 5 min, transferred to a silver impregnation solution (1.5g/L AgNO₃, 0.056% formaldehyde) for 30 min, and then rinsed 3 times with ultrapure water. Image development was carried out with slow agitation for 6–7 min in developer solution (30g/L Na ₂CO₃, 0.056% formaldehyde, 400 µg/L sodium thiosulfate). To stop development and to fix the gel, 10% acetic acid was added directly to the developing solution and incubated with shaking for 3 min. The gel was then rinsed briefly in ultrapure water and dried at room temperature.

Interesting polymorphic bands were excised from gels, rehydrated with 100 µL of double-distilled water for 6 h at 4 °C, and stirred frequently. Tubes were centrifuged at 10,000g for 10 min, and the supernatant transferred into a fresh tube. Aliquots of 5 µL were used as template for re-amplification by PCR in a 50 µL reaction volume. All PCR reactions were carried out with the same EcoRI+2/3 and Msp/Hpa+3 primer combinations used in selective amplification step with the following profile: 94 °C for 1 min, 30 cycles of denaturation at 94 °C for 1 min, annealing at 55 °C for 1 min, and extension at 72 °C for 1 min, ending with a 10-min extension step at 72 °C.

An aliquot of the re-amplified DNA was cloned into a pCR4-TOPO vector using the TOPO TA cloning kit for sequencing (Invitrogen). Ten plasmid DNAs for each transformation were purified from 5 mL of overnight cultures of Escherichia coli in LB medium using the GenElute Plasmid miniprep kit (Sigma). The sequences of both strands of each plasmid were determined after running sequencing reactions (obtained with BigDye^® Terminator v3.1 Cycle Sequencing Kit, Applied Biosystems) on an ABI 3130xl Genetic Analyzer sequencer (Applied Biosystems).

Genome-wide mapping of differentially methylated fragments

The sequences were aligned along two Brassica genomes, as well as the Arabidopsis genome, using different BLAST based approaches [47]. In particular, a BLASTn analysis was used to compare the 24 sequenced bands to the two assembled Brassica genomes B. rapa (Chiifu-401, version 1.1) and B. oleracea (version 2011-08-02-BGI), and also for a comparison with the A. thaliana (version TAIR10 [46]) genome. A tBLASTx analysis was also performed with the B. rapa and A. thaliana genome sequences in order to identify similarities in protein-coding regions. This type of BLAST translates both the query and the subject sequence in all possible reading frames. In order to confirm the presence of the methylated fragments along the Brassicaceae genomes, we also performed a BLASTn alignment versus The Sequence Read Archive (SRA) [74], where the raw sequence data from B. rapa and its subspecies B. rapa subsp. Pekinensis (Chinese cabbage) are stored (http://www.ncbi.nlm.nih.gov/Traces/sra/sra.cgi). Finally for each sequence that showed an alignment to the B. rapa chromosome, the presence of genes both in the alignment region and in the flaking ones were checked, using the Brassica database annotation [45].

In order to enrich the functional description we compiled a Gene Ontology annotation, using the BLAST2GO at the AmiGO web site [75] and setting a cut-off e-value of 0.001.

RNA isolation and cDNA synthesis

Total RNA was isolated from roots and shoots samples using the GenElute Mammalian Total RNA miniprep Kit (Sigma), following the manufacturer’s instructions. Samples were then treated with RQ1 RNase-free DNase (Promega, Madison, WI) following the manufacturer’s instructions. One µg of total RNA was retro-transcribed using random hexamers with the Superscript III Reverse Transcriptase (Invitrogen) according to the manufacturer’s instructions. The PCR reactions using primers for the two housekeeping genes [76] of Brassica rapa, GAPDH (Accession no. AF536826; Forward 5’-CCGCTAACTGCCTTGCTCCACTT-3’, Reverse 5’-GCGGCTCTTCCACCTCTCCAGT-3’) and EF-1-α (Accession no. GO479260; Forward 5’-ATACCAGGCTTGAGCATACCG-3’, Reverse 5’-GCCAAAGAGGCCATCAGACAA-3’), were performed on each sample to check the successful of retro-transcription and the absence of DNA contamination.

Quantitative RT-PCR

All RT-qPCR analyses were performed using an Mx3000P QPCR (Stratagene, La Jolla, CA) system with the SYBR green PCR Master Mix reagent (Applera, Foster City, CA). Using Primer3 software [77] specific primers were designed within the sequences that encode LACERATA (LCR) (Forward 5’-CCGGTACGTATCAGACATGC-3’, Reverse 5’-TCAAAGCGAGTTTTGGGAAT-3’) and trehalose-6-phosphatase synthase S4 (TPS4) (Forward 5’-GCTGATCGTTGGCGTTGA-3’, Reverse 5’-ATGGATTCGCTGCGGAAG-3’). Amplicons of 115 bp and 68 bp were obtained for LCR and TPS4 genes, respectively. The PCR fragments were analysed using a dissociation protocol to ensure that each amplicon was a single product. Amplicons were also sequenced to verify the specificities of the targets. The amplification efficiency was calculated from raw data using the LingRegPCR software [78].

All RT-qPCRs were performed using three biological replicates in a final volume of 25 µl containing 5 µL of cDNA template (previously diluted 1:10), 0.2 µM of each primer, and 12.5 µl of 2× SYBR Green PCR Master Mix (Sigma) according to the manufacturer’s instructions. The following thermal cycling profile was used: 95 °C for 10 min, followed by 50 cycles of 95 °C for 10 s, 57 °C for 15 s, and 72 °C for 15 s. Following cycling, the melting curve was determined in the range 57–95 °C, with a temperature increment of 0.01 °C/sec. Each reaction was run in triplicate (technical replicates). Negative controls included in each run were a reaction conducted in absence of reverse transcriptase and a reaction with no template (2 µL of nuclease-free water instead of 2 µL of cDNA). For negative controls, no signals were observed (data not shown).

Raw Ct data from the MX3000P instrument were exported to a data file and analysed using GeneEx Pro software [79]. During the pre-processing phase, data were corrected for PCR efficiencies and the three technical repeats were averaged. The selected reference genes, GAPDH (Accession no. AF536826) and EF-1-α (Accession no. GO479260) were subsequently used to normalize Ct values [80-82], and quantities were calculated relative to the maximum Ct value. Because our interest was in fold changes in gene expression between groups, we ultimately converted quantities to a logarithmic scale using a log base 2 conversions, which also allowed us to test the normal distribution of values [83].

Extraction of total soluble sugars and trehalose quantification

Total soluble sugars were extracted in 80% ethanol. Because of a lack of plant material, sugar analysis was performed in leaves of Exagone and in the roots of Toccata, only. The ethanol fraction was evaporated under vacuum to dryness (Speed-vac) and resuspended in a small fixed volume of HPLC- water (Acros-Organics): 250 µL for the samples derived from the roots and 400 µL for the samples derived from the shoots. Samples were filtered through a 0.2 µm filter (Acrodisc LC13mm-PVDF; Merck-Eurolab). Sugars were separated on an Aminex HPX-87C resin-based column (BIO-RAD: 125-0095) in degassed and prewarmed (70°C) HPLC-water (Acros-Organics) with a flow-rate of 0.6 mL min-1. The column was protected by a precolumn (Micro-guard cartridges, BIO-RAD: 125-0128) and a filter (Filter Services MN724288) and was kept at 80 °C. Detection was performed through a differential refractometer. Standard curves were generated using trehalose obtained commercially (Sigma) and used at five concentrations prepared in HPLC-water (Acros-Organics); the detector response of the standards was linear in this concentration range and the recovery percentages exceeded 93%. Data analysed by the differential refractometer were collected and analysed with the ‘Valuechrom’ chromatography software (BIO-RAD).

Bisulfite sequencing

Five hundreds nanograms of total DNA extracted from unstressed and salinity stressed shoots and roots of Exagone and Toccata, collected at 7 and 14 DAS, were modified by sodium bisulfite using EZ DNA Methylation-Gold kit (Zymo Research, USA) following the manufacturer’s protocol. An aliquot of 2 µL of bisulfite-treated DNA was used for each PCR reaction (25 µL) using XXX Taq (Life technologies, USA). The primers used for LACERATA locus were: 5’-TTGGTTTAATTGAAAATTGTGAGAGA-3’ (LCR-for) and 5’-TCTCCATCGAAATTAAAAATCCC-3’ (LCR-rev). Primers for TPS4 locus were: 5’-ATTGGGATAGAATTTGAGAGGT-3’ (TPS4-for) and 5’-TTCACTACGAAAATCCCTTT-3’ (TPS4-rev). Amplified products were ligated into the TOPO TA Cloning vector using the TOPO TA Cloning kit for sequencing (Invitrogen). Colonies were grown over night and amplified with specific primer to test the presence of inserts. Plasmids from fifteen positive colonies for each transformation were purified and sequenced. Sequence reactions were performed using Big Dye Terminator v3.1 (Applied Biosystems, Foster City. CA) and run on a ABI 3130xl DNA Analyzer (Applied Biosystems).

Methylation status of DNA was obtained by comparing the sequence of bisulfite-treated DNA with that of untreated DNA, where conversion of a C to T indicated non-methylated C. In contrast, the absence of C to T conversion indicated methylation. Cytosine methylation status in the top strand of obtained nucleotide sequences were calculated using CyMATE v.2 [84]. The methylation level for each of the three kinds of cytosines, CG, CHG, and CHH (where H stands for A, T, or C) was calculated using the following formula: methylated cytosine (%) = [Number of non-converted (methylated) cytosines /Total number of cytosines of each type] x 100.

Statistical analyses

The chi-square was used to test the independence between methylation level and salt stress condition, using SAS Version 9.2 (SAS Institute, Cary, NC). Student’s t-test was also performed using SAS software.”